Robust and Efficient Location of T-Wave Ends in Electrocardiogram (ECG)

Download, the package of Matlab files for T-wave end location and the PhysioNet QT database converted into text format.

package is registered at APP (software protection agency,
under the number IDDN.FR.001.430041.000.S.P.2005.000.31230.

Content of the Readme file

This package contains a T-wave end location program in Matlab and ECG data.

twaveloc: Matlab programs for R peak detection and T-wave end location, written by Qinghua Zhang

qtdata: PhysioNet QT database converted into text format. The original data in binary format
        are available at

qtdbconvert.m: Matlab program to convert the original PhysioNet QT database into text format

PI-1744.pdf: technical report related to the Matlab programs included in this package, also published as:

Q. Zhang, A. Illanes Manriquez, C. Médigue, Y. Papelier, and M. Sorine. An algorithm for robust and efficient location of T-wave ends in electrocardiograms. IEEE Trans. on Biomedical Engineering, 53(12):2544-2552, 2006.

The software for R peak detection and T-wave end location (files in the twaveloc directory)
are distributed under the CeCILL license (
in conformance with French law and following the principles of the GNU GPL.

About the original data used to create the data files included here, the PhysioNet web site
( states that they are are free data, and can be transformed and
redistributed under some rules.

The original binary data have been converted into text format in order to help users who
have difficulties to install the tools necessary for the use of the original binary data.
Following the rules stated on the PhysioNet web site, the Matlab program qtdbconvert.m
is included in this package for reproducing the data transformation.

Matlab programs
The most important files are rpeak.m for R peak detection and twaveend.m for
T-wave end location.

To run the demos, the PhysioNet QT database converted into text format
(located in the qtdata directory included in the same package) must be

IMPORTANT: Before running the demo files qtdemos.m otherdemos.m in Matlab,
the file qtdatapath.m must be edited to indicate the absolute path of the
qtdata directory (do not omit the file separator at the end of the path).

To reproduce the results reported in PI-1744.pdf, run the Matlab files blpsample1.m,
blpsample2.m, blprecord1.m and blprecord2.m.

Author: Qinghua Zhang
Copyright 2005 INRIA